Pith. sign in

REVIEW 3 major objections 6 minor 49 references

J-Invariant Volume Shuffle for Self-Supervised Cryo-Electron Tomogram Denoising on Single Noisy Volume

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A U-shaped blind-spot network with volume-unshuffle can denoise a single noisy cryo-ET volume without paired clean data, outperforming existing self-supervised methods on simulated and real tomograms.

desk verdict A genuine architectural idea, but the gain over SC-Net comes from the guide losses, not from J-invariance. read the letter →

arxiv 2411.15248 v3 pith:VI67VFJX submitted 2024-11-22 eess.IV cs.CV

Xiwei Liu , Mohamad Kassab , Min Xu , Qirong Ho This is my paper · ORCID
classification eess.IVcs.CV
keywords cryo-electrontomographyself-superviseddenoisingblind-spotnetworkJ-invariancevolume-unshuffledilatedchannelattentionsinglenoisyvolumeFSCresolution
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's aim is to show that a cryo-electron tomogram can be denoised from the single noisy volume itself, without paired clean data, noise models, or GAN training. It builds a U-shaped blind-spot network whose predictions for each voxel are made from surrounding voxels only, and it introduces volume-unshuffle/shuffle as a downsampling/upsampling operation that preserves this property through a U-Net's multi-scale path. The central technical claim is that choosing the shuffle volume $v=3$ equal to the dilation factor $d=3$ keeps the network J-invariant, so the model cannot learn to copy the noisy input. On simulated tomograms with different noise types and missing-wedge levels, and on four real cryo-ET datasets, the method reports better denoising and structure preservation than the single-volume self-supervised baselines SC-Net and NMSG, along with better FSC-based resolution estimates. If correct, it gives structural biologists a way to clean the tomogram they already have rather than collecting extra data or training on noise pairs.

What carries the argument

The load-bearing mechanism is volume-unshuffle/shuffle matched to the network's dilation. Volume-unshuffle turns a $D \times H \times W \times C$ volume into $D/v \times H/v \times W/v \times (C \cdot v^3)$ by packing $v \times v \times v$ spatial neighbourhoods into channels; volume-shuffle reverses it. The paper's invariant-preservation condition is $v=d=3$: when the shuffle cell has the same size as the dilation factor of the blind-spot convolutions, a voxel's own value stays out of the receptive field after downsampling, so the U-Net can be deep without learning the identity. Supporting it are sparse centrally masked $5 \times 5 \times 5$ convolutions (central $3 \times 3 \times 3$ masked), dilated channel attention blocks whose depthwise convolution also uses dilation 3, and an edge-representation enhancer that guides training with filtered versions of the input.

What would settle it

Apply volume-unshuffle, the sparse centrally masked dilated convolution, and volume-shuffle to a small synthetic volume with a single nonzero voxel and check whether the output at that voxel is nonzero; if it is, the claimed $v=d=3$ J-invariance guarantee fails, and a noise-only tomogram training run would then reveal how much of the apparent denoising comes from guide-loss smoothing.

Watch

Extended reading notes

Core claim

The authors claim that J-invariance, the requirement that the denoising function's output at a voxel is independent of the input at that same voxel, can be preserved inside a full U-shaped encoder-decoder by replacing pooling and interpolation with volume-unshuffle and volume-shuffle. Volume-unshuffle rearranges a tensor of size $D \times H \times W \times C$ into $D/v \times H/v \times W/v \times (C \cdot v^3)$, moving local spatial neighborhoods into channels; with $v=3$, this aligns with the dilation factor $d=3$ of the centrally masked convolutions and prevents the masked voxel from re-entering the prediction through downsampling. The resulting network uses sparse centrally masked $5 \times 5 \times 5$ convolutions, dilated channel attention blocks, and a loss that combines reconstruction, Gaussian-smoothed contrast guidance, edge enhancement, and total variation. On simulated SHREC2020, SHREC2021, and PolNet volumes and on G. hansenii, Vesicle, Phage T4, and Centriole tomograms, the paper reports higher PSNR/SSIM and lower FSC-based resolution values than SC-Net and NMSG, the main single-volume self-supervised baselines.

Load-bearing premise

The method assumes that, in the reconstructed 3D volume, noise at each voxel has no systematic relationship with its neighbours (so the neighbouring values carry no information about that voxel's noise); the paper's Lemma 1 only shows Gaussian projection noise remains Gaussian in 3D, and real backprojection and interpolation correlate the noise.

Editorial extensions

If this is right

  • Cryo-ET labs can denoise the single reconstructed volume they already have, avoiding the need for paired clean ground truth or pre-trained noise models.
  • Because the U-shape path is preserved under J-invariance, self-supervised denoising gains multi-scale context and long-range dependencies instead of being limited to local masked patches.
  • On real tomograms the method reports lower FSC-based resolution numbers, meaning finer structures may become visible in downstream subtomogram analysis.
  • The $v=d$ condition gives a concrete design rule: for a blind-spot network with dilation $d$, a volume-unshuffle cell of the same size keeps the network honest under downsampling.
  • The same architecture can be applied to any volumetric imaging modality where per-voxel noise is approximately conditionally independent, not only cryo-ET.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Extension: if the $v=d$ lattice alignment is the real mechanism, other matched pairs such as $v=2,d=2$ or $v=4,d=4$ should preserve J-invariance too; testing them would isolate the design rule from quirks of the $5\times5\times5$ mask.
  • Extension: the guide losses push the output toward Gaussian-smoothed and bilaterally filtered versions of the noisy input, so on correlated real noise the reported gains may partly reflect targeted smoothing rather than unbiased signal recovery; a noise-only tomogram experiment could separate the two.
  • Extension: the same volume-unshuffle trick could support self-supervised deconvolution or inpainting of 3D volumes wherever a blind-spot constraint is required, not just denoising.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes a self-supervised denoising method for cryo-electron tomography (cryo-ET) volumes that operates on a single noisy volume. The architecture is a U-shape blind-spot network combining sparse centrally masked convolutions, dilated channel attention (DCA) blocks, and a volume-unshuffle/shuffle (VU/S) down/up-sampling scheme that the authors claim preserves J-invariance. Training uses a weighted sum of four losses: a reconstruction loss against the noisy input, a contrast guidance loss toward a Gaussian-filtered version of the input, an edge enhancement loss toward a bilateral-filtered version, and a total variation term. The method is evaluated on three simulated datasets (SHREC2020, SHREC2021, PolNet) with additive noise, on simulated missing-wedge conditions, and on four real cryo-ET datasets, reporting PSNR/SSIM and FSC-e/o resolution improvements over SC-Net and NMSG.

Significance. If the reported gains were attributable to the proposed J-invariant volume-shuffle architecture, the paper would make a useful contribution to cryo-ET denoising, where single-volume self-supervised methods are practically important. The paper provides a code link and a broad evaluation across simulated and real datasets, which is a strength. However, the ablation study in Table 5 directly undercuts the attribution of the gains to the J-invariant architecture: removing the auxiliary filtered-target losses (Lguide, Ledge) reduces performance to near or below the SC-Net baseline. The central theoretical claim about J-invariance preservation under VU/S is only deferred to a missing supplement, and the formal equations defining VU/S appear garbled. As a result, the empirical superiority of the full method is clear, but the scientific claim about the mechanism is not supported by the evidence presented.

major comments (3)
  1. [Section 4.2, Eqs. (2)-(6) and Supplementary S2] The formal definition of volume-unshuffle is garbled and the proof that it preserves J-invariance is only in the missing supplement. For example, Eq. (3) writes i' = v * floor(i / v3) + (i mod v), which mixes v and v3 in a way that is not consistent with standard pixel-shuffle indexing, and Eq. (6) defines m' with an expression that is not dimensionally coherent. Since the claim that VU/S with v = d = 3 preserves J-invariance is the paper's main novelty, the mapping must be defined with correct, self-contained equations and the preservation proof must appear in the main text or in an actually included supplement.
  2. [Section 4.3, Eq. (11) and Table 5] The self-supervised unbiasedness argument from Noise2Self justifies only Lrec (Eq. 7). Lguide (Eq. 8) and Ledge (Eq. 9) use deterministic filtered versions of the same noisy volume as regression targets; these targets are biased estimates of the clean signal, so the J-invariance theory does not justify the full objective. Table 5 quantifies the consequence: removing Ledge drops PSNR on SHREC21 from 37.62 to 30.71, and removing all auxiliary losses ('w/o all') gives 28.99, which is close to SC-Net's 29.78. Thus the reported margin over SC-Net is produced mostly by the hand-crafted filtered targets rather than by the J-invariant volume-shuffle architecture, undermining the abstract's claim that the method 'achieves superior performance' in the sense of a generally applicable self-supervised mechanism.
  3. [Section 3, Lemma 1 and Eq. (1)] The guarantee that a J-invariant network trained with a self-supervised loss is unbiased toward the clean signal requires the noise at each voxel to have zero conditional mean given the surrounding voxels, or at least approximate voxel-wise independence. Lemma 1 only states that additive Gaussian noise in 2D projections remains Gaussian in the 3D reconstruction; it does not establish independence or conditional unbiasedness. Real cryo-ET reconstruction involves backprojection, missing wedge, and interpolation, which correlate the noise. The paper should either prove the needed conditional property under the stated assumptions or explicitly state it as an additional assumption and discuss its plausibility in the cryo-ET setting.
minor comments (6)
  1. [Section 4.2, Eq. (2)-(6)] The sentence 'employing volume-unshuffle with a volume size v = 3 equals to dilation factor d = 3' is grammatically awkward and should clarify why the equality v = d is necessary for J-invariance.
  2. [Section 4.3] There are typos: 'consrast' should be 'contrast', and 'biliteral' should be 'bilateral'.
  3. [Section 5.1] The phrase '1083 patch size' presumably means 108^3; please write it in standard mathematical notation.
  4. [Table 7 and its caption] The caption contains a duplicated word: 'D-BSN block replaces replaces the DCA module' should read 'replaces the DCA module'.
  5. [Section 5.1] 'NVDIA GeForce RTX 4090' should be 'NVIDIA GeForce RTX 4090'.
  6. [General] Several supplementary sections (S1, S2, S3, S6, S7) are referenced but not included in the submitted manuscript; without them the reader cannot verify the central proof, the dataset details, or the visual comparisons.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is an empirical architecture with self-supervised losses on the same noisy volume; no claimed prediction reduces by construction to its inputs.

full rationale

The paper's central claims are (i) a U-shape blind-spot network with sparse centrally masked convolutions, DCA blocks, and volume-unshuffle/shuffle can denoise a single noisy cryo-ET volume, and (ii) volume-unshuffle/shuffle preserves J-invariance so the Noise2Self-style reconstruction loss remains unbiased. Neither claim is definitionally circular: J-invariance is defined independently of volume-unshuffle/shuffle (Definition 1), and the statement that volume-unshuffle/shuffle with v = d = 3 preserves J-invariance is a mathematical property to be established, not an assumed equivalence. The loss in Eq. 11 is a weighted sum of Lrec (regression to the noisy input, standard Noise2Self), Lguide (regression to a Gaussian-filtered version of the same input), Ledge (edge-map distance to a bilateral-filtered version), and LTV. These auxiliary terms are explicit design choices, not fitted parameters disguised as predictions, and the full output is a compromise among them rather than identical to any single filtered input. Table 5 shows that removing the auxiliary losses degrades PSNR, which attributes part of the gain to the filtering guides, but that is an empirical attribution issue, not a circular derivation. The comparison against external baselines on simulated data with ground truth (Table 2) and real data with FSCe/o (Table 4) gives the central claim independent empirical content. The missing or garbled supplementary proof of volume-unshuffle/shuffle J-invariance is a completeness and correctness concern, not circularity. No load-bearing self-citation chain is present: the cited prior works (e.g., Noise2Self, D-BSN, PUCA) are external and are used as inspiration or baselines, not as the sole justification of the paper's results.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The artifact ledger shows the method is built on standard self-supervised theory plus a custom J-invariance claim deferred to the supplement. The hand-set loss weights are free parameters with demonstrated sensitivity in Table 5. No new physical entities are postulated.

free parameters (5)
  • lambda1 = 0.8
    Reconstruction loss weight in Eq. 11; set by hand without stated criterion.
  • lambda2 = 0.5
    Contrast guidance loss weight; set by hand; ablation shows strong influence on PSNR/SSIM.
  • lambda3 = 0.05
    Edge enhancement loss weight; set by hand.
  • lambda4 = 0.01
    Total variation loss weight; set by hand.
  • volume shuffle size v = 3
    Chosen to equal the dilation factor d=3 for J-invariance preservation; a design choice rather than a data fit, but it constrains the architecture.
assumptions (4)
  • domain assumption Lemma 1: additive Gaussian noise in 2D projections remains Gaussian in the 3D reconstruction (V = Psi + N).
    Section 3, Eq. 1; standard assumption for cryo-ET linear reconstruction, inherited from Yang et al. [42].
  • standard math Noise2Self theorem: a J-invariant function minimizes the self-supervised loss toward the conditional expectation, hence denoises.
    Invoked in Section 3, Definition 1; background ML theory from Batson and Royer [4].
  • domain assumption Voxel-wise noise independence (or zero conditional mean) in the 3D volume, required for the blind-spot training target to be unbiased.
    Unstated but load-bearing; cryo-ET reconstruction correlates noise, so this is a fragile premise.
  • ad hoc to paper Volume-unshuffle with v=3 equal to dilation factor d=3 preserves J-invariance of the whole network.
    Section 4.2; proof deferred to Supplementary S2, equations 2-6 are garbled in main text.

how reviews work

0 comments
Cite this review

Pith. "Pith review of J-Invariant Volume Shuffle for Self-Supervised Cryo-Electron Tomogram Denoising on Single Noisy Volume." pith.science (2026). https://pith.science/paper/VI67VFJX

@misc{pith2026241115248,
  author       = {Pith},
  title        = {Pith review of: J-Invariant Volume Shuffle for Self-Supervised Cryo-Electron Tomogram Denoising on Single Noisy Volume},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VI67VFJX}},
  note         = {Machine review of arXiv:2411.15248}
}
read the original abstract

Cryo-Electron Tomography (Cryo-ET) enables detailed 3D visualization of cellular structures in near-native states but suffers from low signal-to-noise ratio due to imaging constraints. Traditional denoising methods and supervised learning approaches often struggle with complex noise patterns and the lack of paired datasets. Self-supervised methods, which utilize noisy input itself as a target, have been studied; however, existing Cryo-ET self-supervised denoising methods face significant challenges due to losing information during training and the learned incomplete noise patterns. In this paper, we propose a novel self-supervised learning model that denoises Cryo-ET volumetric images using a single noisy volume. Our method features a U-shape J-invariant blind spot network with sparse centrally masked convolutions, dilated channel attention blocks, and volume unshuffle/shuffle technique. The volume-unshuffle/shuffle technique expands receptive fields and utilizes multi-scale representations, significantly improving noise reduction and structural preservation. Experimental results demonstrate that our approach achieves superior performance compared to existing methods, advancing Cryo-ET data processing for structural biology research

Figures

Figures reproduced from arXiv: 2411.15248 by the authors.

Figure 1
Figure 1. The architecture of proposed method. The noisy input volume is first preprocessed using (i) 3D Gaussian filters to generate a [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Volume unshuffle and shuffle. To address this issue, we introduce the volume-unshuffle technique for volumetric tomogram images. Volume￾unshuffle transforms a tensor of size D \times H \times W \times C into a reshaped tensor of size \frac {D}{v} \times \frac {H}{v} \times \frac {W}{v} \times (C \cdot v^3) , where v represents the shuffle volume size, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Details of edge representation enhancer. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Visual results of the real data. severe MW, the model still preserves structure well, demon￾strating its robustness. 5.3. Real-world Datasets We evaluate on four real cryo-ET datasets: G. hansenii bio9-2, Vesicle, Escherichia phage T4, and Centriole. The G. hansenii bi…
Figure 5
Figure 5. Figure 5: Examples of FSCe/o curves for the G.hanseni and Phage. Red dash line in the figures point out the position of FSCe/o=0.5. Quantitative analysis. As ground-truth is unavailable in real-world situations, we adopt a cross-validation metric [PITH_FULL_IMAGE:figures/full_f…
Figure 6
Figure 6. Figure 6: Visual results of component analysis on loss functions. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 33 canonical work pages

  1. [1]

    Cryo-electron microscopy of viruses.Na- ture, 308(5954):32–36, 1984

    Marc Adrian, Jacques Dubochet, Jean Lepault, and Alas- dair W McDowall. Cryo-electron microscopy of viruses.Na- ture, 308(5954):32–36, 1984. 1

    work page 1984

  2. [2]

    Adaptive multi-column deep neural networks with applica- tion to robust image denoising

    Forest Agostinelli, Michael R Anderson, and Honglak Lee. Adaptive multi-column deep neural networks with applica- tion to robust image denoising. Advances in neural informa- tion processing systems, 26, 2013. 2

    work page 2013

  3. [3]

    Layer normalization

    Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hin- ton. Layer normalization. arXiv preprint arXiv:1607.06450,

    arXiv

  4. [4]

    Noise2self: Blind denoising by self-supervision

    Joshua Batson and Loic Royer. Noise2self: Blind denoising by self-supervision. In International Conference on Machine Learning, pages 524–533. PMLR, 2019. 1, 3

    2019

  5. [5]

    Topaz-denoise: general deep denoising models for cryoem and cryoet

    Tristan Bepler, Kotaro Kelley, Alex J Noble, and Bonnie Berger. Topaz-denoise: general deep denoising models for cryoem and cryoet. Nature communications , 11(1):5208,

    work page

  6. [6]

    A non-local algorithm for image denoising

    Antoni Buades, Bartomeu Coll, and J-M Morel. A non-local algorithm for image denoising. In 2005 IEEE computer so- ciety conference on computer vision and pattern recognition (CVPR’05), volume 2, pages 60–65. Ieee, 2005. 1, 2, 6

    work page 2005

  7. [7]

    Content-aware image restoration for electron microscopy

    Tim-Oliver Buchholz, Alexander Krull, R ´eza Shahidi, Gaia Pigino, G´asp´ar J´ekely, and Florian Jug. Content-aware image restoration for electron microscopy. Methods in cell biology, 152:277–289, 2019. 2

    work page 2019

  8. [8]

    A resolution criterion for electron tomography based on cross-validation

    Giovanni Cardone, Kay Gr ¨unewald, and Alasdair C Steven. A resolution criterion for electron tomography based on cross-validation. Journal of structural biology, 151(2):117– 129, 2005. 8

    work page 2005

Show all 49 references
  1. [9]

    Simple baselines for image restoration

    Liangyu Chen, Xiaojie Chu, Xiangyu Zhang, and Jian Sun. Simple baselines for image restoration. In European confer- ence on computer vision , pages 17–33. Springer, 2022. 2, 5

    2022

  2. [10]

    Rethinking coarse-to-fine approach in sin- gle image deblurring

    Sung-Jin Cho, Seo-Won Ji, Jun-Pyo Hong, Seung-Won Jung, and Sung-Jea Ko. Rethinking coarse-to-fine approach in sin- gle image deblurring. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 4641–4650,

  3. [11]

    Adam: A method for stochastic opti- mization

    P Kingma Diederik. Adam: A method for stochastic opti- mization. (No Title), 2014. 6

    2014

  4. [12]

    Denoising prior driven deep neural network for image restoration

    Weisheng Dong, Peiyao Wang, Wotao Yin, Guangming Shi, Fangfang Wu, and Xiaotong Lu. Denoising prior driven deep neural network for image restoration. IEEE transactions on pattern analysis and machine intelligence , 41(10):2305– 2318, 2018. 2

    2018

  5. [13]

    Noise reduc- tion in electron tomographic reconstructions using nonlin- ear anisotropic diffusion

    Achilleas S Frangakis and Reiner Hegerl. Noise reduc- tion in electron tomographic reconstructions using nonlin- ear anisotropic diffusion. Journal of structural biology , 135(3):239–250, 2001. 1, 2, 6

    2001

  6. [14]

    Three-dimensional tomographic reconstruction in high voltage electron microscopy

    J Frank, BF McEwen, M Radermacher, JN Turner, and CL Rieder. Three-dimensional tomographic reconstruction in high voltage electron microscopy. Journal of electron mi- croscopy technique, 6(2):193–205, 1987. 1

    1987

  7. [15]

    Spider and web: processing and visualization of images in 3d electron microscopy and related fields

    Joachim Frank, Michael Radermacher, Pawel Penczek, Jun Zhu, Yanhong Li, Mahieddine Ladjadj, and Ardean Leith. Spider and web: processing and visualization of images in 3d electron microscopy and related fields. Journal of structural biology, 116(1):190–199, 1996. 6

    1996

  8. [16]

    Improved training of wasserstein gans

    Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky, Vincent Dumoulin, and Aaron C Courville. Improved training of wasserstein gans. Advances in neural information processing systems, 30, 2017. 1

    2017

  9. [17]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 5

    2016

  10. [18]

    Image-to-image translation with conditional adver- sarial networks

    Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A Efros. Image-to-image translation with conditional adver- sarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1125–1134,

  11. [19]

    Natural image denoising with convolutional networks

    Viren Jain and Sebastian Seung. Natural image denoising with convolutional networks. Advances in neural informa- tion processing systems, 21, 2008. 1, 2

    2008

  12. [20]

    Puca: patch-unshuffle and channel attention for enhanced self-supervised image denoising

    Hyemi Jang, Junsung Park, Dahuin Jung, Jaihyun Lew, Ho Bae, and Sungroh Yoon. Puca: patch-unshuffle and channel attention for enhanced self-supervised image denoising. Ad- vances in Neural Information Processing Systems, 36, 2024. 3, 4

    2024

  13. [21]

    Computer visualization of three-dimensional image data using imod

    James R Kremer, David N Mastronarde, and J Richard McIn- tosh. Computer visualization of three-dimensional image data using imod. Journal of structural biology , 116(1):71– 76, 1996. 7

    1996

  14. [22]

    Noise2void-learning denoising from single noisy images

    Alexander Krull, Tim-Oliver Buchholz, and Florian Jug. Noise2void-learning denoising from single noisy images. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 2129–2137, 2019. 1, 3, 6

    2019

  15. [23]

    High-quality self-supervised deep image denoising

    Samuli Laine, Tero Karras, Jaakko Lehtinen, and Timo Aila. High-quality self-supervised deep image denoising. Ad- vances in Neural Information Processing Systems, 32, 2019. 3

    2019

  16. [24]

    Ap- bsn: Self-supervised denoising for real-world images via asymmetric pd and blind-spot network

    Wooseok Lee, Sanghyun Son, and Kyoung Mu Lee. Ap- bsn: Self-supervised denoising for real-world images via asymmetric pd and blind-spot network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17725–17734, 2022. 1, 2, 3

    2022

  17. [25]

    Noise2noise: Learning image restoration without clean data

    Jaakko Lehtinen, Jacob Munkberg, Jon Hasselgren, Samuli Laine, Tero Karras, Miika Aittala, and Timo Aila. Noise2noise: Learning image restoration without clean data. arXiv preprint arXiv:1803.04189, 2018. 2

    2018 arXiv

  18. [26]

    Nonlocal transform-domain filter for volumetric data denoising and reconstruction

    Matteo Maggioni, Vladimir Katkovnik, Karen Egiazarian, and Alessandro Foi. Nonlocal transform-domain filter for volumetric data denoising and reconstruction. IEEE trans- actions on image processing, 22(1):119–133, 2012. 2, 6

    2012

  19. [27]

    Improving the quality of electron tomography image volumes using pre-reconstruction filter- ing

    Mauro Maiorca, Eric Hanssen, Edmund Kazmierczak, Bo- humil Maco, Misha Kudryashev, Richard Hall, Harry Quiney, and Leann Tilley. Improving the quality of electron tomography image volumes using pre-reconstruction filter- ing. Journal of structural biology , 180(1):132–142, 2012. 2

    2012

  20. [28]

    Image restoration using very deep convolutional encoder-decoder networks with symmetric skip connections.Advances in neu- ral information processing systems, 29, 2016

    Xiaojiao Mao, Chunhua Shen, and Yu-Bin Yang. Image restoration using very deep convolutional encoder-decoder networks with symmetric skip connections.Advances in neu- ral information processing systems, 29, 2016. 2

    2016

  21. [29]

    Simulating the cellular context in synthetic datasets for cryo-electron tomography

    Antonio Martinez-Sanchez, Lorenz Lamm, Marion Jasnin, and Harold Phelippeau. Simulating the cellular context in synthetic datasets for cryo-electron tomography. bioRxiv, pages 2023–05, 2023. 6

    2023

  22. [30]

    V-net: Fully convolutional neural networks for volumetric medical image segmentation

    Fausto Milletari, Nassir Navab, and Seyed-Ahmad Ahmadi. V-net: Fully convolutional neural networks for volumetric medical image segmentation. In 2016 fourth international conference on 3D vision (3DV), pages 565–571. Ieee, 2016. 6

    2016

  23. [31]

    Noisier2noise: Learning to denoise from unpaired noisy data

    Nick Moran, Dan Schmidt, Yu Zhong, and Patrick Coady. Noisier2noise: Learning to denoise from unpaired noisy data. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , pages 12064–12072,

  24. [32]

    Stripe and ring artifact removal with combined wavelet—fourier filtering

    Beat M ¨unch, Pavel Trtik, Federica Marone, and Marco Stam- panoni. Stripe and ring artifact removal with combined wavelet—fourier filtering. Optics express , 17(10):8567– 8591, 2009. 1

    2009

  25. [33]

    Fourier slice photography

    Ren Ng. Fourier slice photography. In ACM Siggraph 2005 Papers, pages 735–744. 2005. 2

    2005

  26. [34]

    Pytorch: An im- perative style, high-performance deep learning library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An im- perative style, high-performance deep learning library. Ad- vances in neural information processing systems ...

    2019

  27. [35]

    U- net: Convolutional networks for biomedical image segmen- tation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- net: Convolutional networks for biomedical image segmen- tation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, pa...

    2015

  28. [36]

    Bilateral filtering for gray and color images

    Carlo Tomasi and Roberto Manduchi. Bilateral filtering for gray and color images. In Sixth international conference on computer vision (IEEE Cat. No. 98CH36271) , pages 839–

  29. [37]

    The promise and the challenges of cryo-electron tomography

    Martin Turk and Wolfgang Baumeister. The promise and the challenges of cryo-electron tomography. FEBS letters, 594(20):3243–3261, 2020. 1

    2020

  30. [38]

    Efficient automatic noise re- duction of electron tomographic reconstructions based on iterative median filtering

    Peter van der Heide, Xiao-Ping Xu, Brad J Marsh, Dorit Hanein, and Niels V olkmann. Efficient automatic noise re- duction of electron tomographic reconstructions based on iterative median filtering. Journal of structural biology , 158(2):196–204, 2007. 2

    2007

  31. [39]

    Unpaired learning of deep image denoising

    Xiaohe Wu, Ming Liu, Yue Cao, Dongwei Ren, and Wang- meng Zuo. Unpaired learning of deep image denoising. In European conference on computer vision , pages 352–368. Springer, 2020. 2, 3, 5

    2020

  32. [40]

    Image denois- ing and inpainting with deep neural networks

    Junyuan Xie, Linli Xu, and Enhong Chen. Image denois- ing and inpainting with deep neural networks. Advances in neural information processing systems, 25, 2012. 1, 2

    2012

  33. [41]

    Self-supervised noise modeling and sparsity guided electron tomography volumetric image denoising

    Zhidong Yang, Dawei Zang, Hongjia Li, Zhao Zhang, Fa Zhang, and Renmin Han. Self-supervised noise modeling and sparsity guided electron tomography volumetric image denoising. Ultramicroscopy, 255:113860, 2024. 2, 6

    2024

  34. [42]

    Self-supervised cryo-electron tomography volumetric image restoration from single noisy volume with sparsity constraint

    Zhidong Yang, Fa Zhang, and Renmin Han. Self-supervised cryo-electron tomography volumetric image restoration from single noisy volume with sparsity constraint. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 4056–4065, 2021. 1, 3, 6

    2021

  35. [43]

    Dual adversarial network: Toward real-world noise removal and noise generation

    Zongsheng Yue, Qian Zhao, Lei Zhang, and Deyu Meng. Dual adversarial network: Toward real-world noise removal and noise generation. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part X 16, pages 41–58. Springer, 2020. 2

    2020

  36. [44]

    Restormer: Efficient transformer for high-resolution image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Mu- nawar Hayat, Fahad Shahbaz Khan, and Ming-Hsuan Yang. Restormer: Efficient transformer for high-resolution image restoration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5728–5739,

  37. [45]

    Multi-stage progressive image restoration

    Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, Ming-Hsuan Yang, and Ling Shao. Multi-stage progressive image restoration. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14821–14831, 2021. 2

    2021

  38. [46]

    Mm-bsn: Self-supervised image denoising for real-world with multi-mask based on blind-spot network

    Dan Zhang, Fangfang Zhou, Yuwen Jiang, and Zhengming Fu. Mm-bsn: Self-supervised image denoising for real-world with multi-mask based on blind-spot network. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4188–4197, 2023. 1, 2, 3

    2023

  39. [47]

    Unleashing the power of self-supervised im- age denoising: A comprehensive review

    Dan Zhang, Fangfang Zhou, Yuanzhou Wei, Xiao Yang, and Yuan Gu. Unleashing the power of self-supervised im- age denoising: A comprehensive review. arXiv preprint arXiv:2308.00247, 2023. 2

    2023 arXiv

  40. [48]

    Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising

    Kai Zhang, Wangmeng Zuo, Yunjin Chen, Deyu Meng, and Lei Zhang. Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising. IEEE transactions on image processing, 26(7):3142–3155, 2017. 1, 2, 6

    2017

  41. [49]

    Ffdnet: Toward a fast and flexible solution for cnn-based image denoising

    Kai Zhang, Wangmeng Zuo, and Lei Zhang. Ffdnet: Toward a fast and flexible solution for cnn-based image denoising. IEEE Transactions on Image Processing, 27(9):4608–4622,

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.